next up previous contents
Next: 3.11 Common Acquisition/Observation Problems Up: 3.10 Exposure Time Estimates Previous: 3.10.2 Estimates Based on

3.10.3 Other Factors Influencing Exposure Time Estimates

3.10.3.1 Variation of Optimal Exposure with Long-Wavelength Camera

Note that the older long-wavelength data were usually obtained with the LWR camera, but that the LWP is now in use. The low dispersion exposure times may be scaled from one camera to the other, depending upon the wavelength region of interest. The following relations give approximate scaling factors for estimating LWP exposure times (t) from previous LWR (i.e. 5.0 kV configuration) exposures (all large aperture spectra):

t (LWP low disp) = 0.8·t (LWR low disp) for 2800 Å
t (LWP low disp) = 1.1·t (LWR low disp) for 2400 Å
t (LWP high disp) = 0.9·t (LWR high disp) for 2800 Å continuum
t (LWP high disp) = 0.8·t (LWR high disp) for Mg II emission lines

The LWR camera was permanently reconfigured to the 4.5 kv configuration in October 1985. In the 4.5 kv configuration, the LWR camera is 1.37 less sensitive at all wavelengths than for the 5.0 kv configuration. This fact should be taken into account for recent LWR images. Most LWR images taken between October 1983 and 1985 were obtained with the 5.0 kv configuration, but there are exceptions.

3.10.3.2 Variation of Optimal Exposure with Wavelength

The values of E¯¹(Lambda) given in Figure 3.5 were chosen to give a peak camera response of approximately 200 DN. In some parts of the image, exposure levels greater than approximately 200 DN will be processed less accurately, as they exceed the highest level of the Intensity Transfer Function (ITF), the calibration which linearizes the raw data. Table 3.5 gives an estimate of the DN of the highest unsaturated ITF level as a function of wavelength for low-dispersion spectra. A DN value of 255 is saturated and all intensity information is lost. For pixels having DN values greater than the highest unsaturated ITF level but less than 255, as given in Table 3.5 for low dispersion spectra, the given pixel's intensity is derived by extrapolating the data for the lower ITF levels. The flux level for such pixels therefore have reduced photometric accuracy.

 
Wavelength SWP Wavelength LWR LWP
         
1200 Å 197 DN 2100 Å 245 DN 205 DN
1300 216 2300 245 220
1400 235 2500 240 240
1500 245 2700 225 245
1600 243 2900 220 245
1700 239 3100 190 245
1800 228 3300 180 245
1900 242      
Table 3.5: DN Value for Highest Unsaturated ITF Level

3.10.3.3 Non-Target Contributions to the Signal

When deriving exposure time estimates, all contributions to the signal must be included, such as camera phosphorescence (Section 4.6) and radiation fogging (Section 4.9). In addition, the spectrum is recorded on a repeatable camera pedestal. Table 3.6 lists the pedestal levels for the three cameras in the region of maximum sensitivity.

 
Camera Pedestal
   
LWP 38 DN
LWR 23 DN
SWP 25 DN
Table 3.6: Camera Pedestals

3.10.3.4 Overexposing the Camera to Optimize Spectral Regions

Unreddened stars with spectral types B3 and earlier will produce maximum camera response at 1300Å and 2700Å in the short- and long-wavelength spectral ranges, respectively. Cooler and reddened stars will produce maximum camera response at 1800-1900 Å and 2800-3000 Å. The continuum around 1500 Å and 2200 Å will be less heavily exposed, regardless of stellar type. In the echelle (high-dispersion) mode, the intensity decreases toward the ends of each order. Hence, to obtain well-exposed spectra of a feature near the end of an order (for example, Mg II 2803Å), it may be necessary to overexpose the central part of the order. The telescope operations staff should be informed before beginning an exposure which will intentionally overexpose part of the camera.

If there is some uncertainty in the target's UV flux and a heavy overexposure (i.e. 20X or more) is needed to bring up certain spectral features, a test exposure is strongly advised. Very heavy overexposures (i.e. >500X) may permanently damage the cameras. Even at exposure levels which do not damage the cameras, phosphorescence in the UVC can affect subsequent images for up to several days after the overexposure has occurred and thus possibly affect both you and subsequent observers and their programs (see Section 4.6). This phosphorescence can be generated either from a single heavy overexposure, or a series of moderate overexposures (i.e. <2X). Project approval is required for all overexposures of >100X.

3.10.3.5 Examples

Example 1. A previous IUE exposure yielded a peak signal of 85 DN and a background of 45 DN in a 3 hour exposure on the SWP camera. Thus the net signal was accumulating at a rate of 13 DN/hour. (Note that the background above the pedestal of 25 DN was also accumulating at a rate of 7 DN per hour). For a 7 hour exposure, the predicted peak signal would be

25 DN (Pedestal) + 7 x 7 DN/hour (Background Phosphorescence) + 7 x 13 DN/hour (Signal) = 160 DN.

If the background radiation were averaging 1.0 volts for the last hour of the exposure, an additional 10 DN background would be added to the more sensitive regions of the camera and about 7 DN to the remaining regions resulting in a peak signal of 170 DN. If the radiation climbed above 1.0 volt, the GO might shorten the exposure since the background noise from radiation would be accumulating at a rate several times that of the target.

Example 2. A previous LWR low-dispersion spectrum was obtained with optimum DN levels in 30 seconds. The spectrum is that of a continuum point-source. To estimate the exposure time for an LWP, high-dispersion small aperture spectrum, we have

t(needed) = 65 (low to high dispersion) x 2 (large to small aperture) x 0.9 (LWR to LWP high dispersion) x 30 seconds (original exposure)

t(needed) = 58.5 minutes


next up previous contents
Next: 3.11 Common Acquisition/Observation Problems Up: 3.10 Exposure Time Estimates Previous: 3.10.2 Estimates Based on

Last updated: 24 July 1997
jrc